The image of motor units architecture in the mechanomyographic signal during the single motor unit contraction: in vivo and simulation study

Abstract

The mechanomyographic (MMG) signal analysis has been performed during single motor unit (MU) contractions of the rat medial gastrocnemius muscle. The MMG has been recorded as a muscle surface displacement by using a laser distance sensor. The profiles of the MMG signal let to categorize these signals for particular MUs into three classes. Class MMG-P (positive) comprises MUs with the MMG signal similar to the force signal profile, where the distance between the muscle surface and the laser sensor increases with the force increase. The class MMG-N (negative) has also the MMG profile similar to the force profile, however the MMG is inverted in comparison to the force signal and the distance measured by using laser sensor decreases with the force increase. The third class MMG-M (mixed) characterize the MMG which initially increases with the force increases and when the force exceeds some level it starts to decrease towards the negative values. The semi-pennate muscle model has been proposed, enabling estimation of the MMG generated by a single MU depending on its localization. The analysis have shown that in the semi-pennate muscle the localization of the MU and the relative position of the laser distance sensor determine the MMG profile and amplitude. Thus, proposed classification of the MMG recordings is not related to the physiological types of MUs, but only to the MU localization and mentioned sensor position. When the distance sensor is located over the middle of the muscle belly, a part of the muscle fibers have endings near the location of the sensor beam. For the MU MMG of class MMG-N the deflection of the muscle surface proximal to the sensor mainly influences the MMG recording, whereas for the MU MMG class MMG-P, it is mainly the distal muscle surface deformation. For the MU MMG of MMG-M type the effects of deformation within the proximal and distal muscle surfaces overlap. The model has been verified with experimental recordings, and its responses are consistent and adequate in comparison to the experimental data.